Hi guys,
Here's the schematic I use for simulations, with DC voltages.
The filter is passive, and followed by a differential tubestage with a current source in the tail. Current is about 5...6mA per tube. I have a ECC82/12AU7 in there at the moment, but any tube with a mu around 20 can be used. I use the original -12V regulator and circuit on the player's PCB to create an extra -5V for the current source. That makes sure it sees enough voltage to keep working properly, in case the tube is swapped with one that has less bias voltage. Just see to it that there's enough voltage between the grid and kathodes to allow for signal swing.
Next week i'm going to test it and see if the supply voltage can be lowered more. In the end I want to use a tube that's not very hard to get, and performs good at lower B+.
Regards,
Ray.
Here's the schematic I use for simulations, with DC voltages.
The filter is passive, and followed by a differential tubestage with a current source in the tail. Current is about 5...6mA per tube. I have a ECC82/12AU7 in there at the moment, but any tube with a mu around 20 can be used. I use the original -12V regulator and circuit on the player's PCB to create an extra -5V for the current source. That makes sure it sees enough voltage to keep working properly, in case the tube is swapped with one that has less bias voltage. Just see to it that there's enough voltage between the grid and kathodes to allow for signal swing.
Next week i'm going to test it and see if the supply voltage can be lowered more. In the end I want to use a tube that's not very hard to get, and performs good at lower B+.
Regards,
Ray.
6h5c said:
Hi Ray.
Are you still trying to use existing component locations for the filter??
How are you mounting the tube - on a daughterboard? Horizontally or vertically?
Andy
6h5c said:
active(opamp) or passive filter with tube buffer
with passive filter you'll hear more signature of the dac
maybe good or not unknown at this stage.
still using thorsten's buffer or have you modded it too?
allan
ps just picked up tent clock, now to build psu for it😀
poynton said:
post before
I didn't see it!!!!!!!
that'll teach me to refresh screen while i'm doing layout for psu 😀
old program dos mode🙄
if anything running in background it kills my work😡
very frustrating
allan
ps ray very nice
I'm going to have to try my 6922's
low voltage
Yes i would like to put everthing in the cdp case
poynton said:
awpagan said:
I'm not using Thorstens filter. It's more like a 3rd order Bessel now. The group-delay is much more constant as well as the input impedance. The impedance is higher compared to Thorsten's so I guess it will be o.k. for the DAC to drive it.
The filter can be fit on the player's PCB: first there's a R-C section, like in the player, and then a L-CR, which is R-CR in the player. So the original footprints and lay-out can be used 😀.
In the end I want to mount the tubes horizontally, on a piece of right angled aluminium or something like that. I'll put two 22mm holes in it for the tubesockets and hardwire the circuit. There's no need for a PCB. Maybe i'll mount it alongside the PCB, maybe on it, to get short wires to the PCB. Depends a bit on the height.
All I need now is a tranny that supplies the heater voltage and +150...180V. I'm thinking about a custom 36VA with EI66 core that replaces the original tranny.
Regards,
Ray.
6h5c said:I'm not sure either
I don't have a good scope to take a look at the waveform. But technically there could be reflections bouncing back and forward that interfere with the original waveform.
I've done some more tinkering with the passive filters, and did some more simulations. Here's a PDF with the latest i've come up with.
The circuit of the Butterworth filters matches that of the player's PCB, so it's easier to fit the components on the original board. Also the input impedance is much more constant compared to Thorsten's circuit. But I don't know at what impedance the DAC is most happy.
I made filters with various attenuation, because I still have to decide what buffer i'm going to use. Maybe the modified HDAM, or a Lundahl tranny, or a tubestage. Have to do some more experiments.
Regards,
Ray.
http://www.diyaudio.com/forums/attachment.php?s=&postid=851364&stamp=1140533190
Hi Ray.
On your simulations, what is the attenuation of circuits 4,5,6 ?
Andy
poynton said:
Ehm....the number with the "minus" in front of it?
😀
But these are the "old" circuits, with not-so-constant group-delay.
Ray.
My clock currently runs from +20V rail to an Audiocom 12V reg with 2 coils and 2 16V 100uF Black gate FK caps.
Im going to take the supply from the raw ac and regulate it to the Audiocom so its isolated from the +20V which supplies my regs for opamps.
Here is the proposed diagram, any advise would be welcome.
Im going to take the supply from the raw ac and regulate it to the Audiocom so its isolated from the +20V which supplies my regs for opamps.
Here is the proposed diagram, any advise would be welcome.
Attachments
rowemeister said:
for 12volt out of the audiocom reg
i think it may need more than 12volt in?
do a search for info on the LT reg, I read they can be difficult on layout.
allan
ps also how about a inductor (coil with beads) before the super reg.
awpagan said:
Yeah that was an error awpagan. You can only get the LM1086 up to 5V and then you add zener diode to gnd for its ref. It will actually be 15V.
Brent
rowemeister said:
I thought that might be a typo😀
How about a bit of CLC or LC, between regs
problem voltage drop?
allan
6h5c said:
old age moment !!!!!!I realise that you have revised the circuit...my problem is that I cannot get hold of 10.0uH chokes !
Any chance you could compute the values for me with 4.7uH ????
Andy
rowemeister said:
Hi Brent,
Why don't you use the adjustable version?
I don't like the idea of lifting the regs reference with a noisy zener
. This will be transferred to the output.It will also enable you to decouple the adjust-pin and get better rejection ratio
.Why do you use two 3300uF BG's and two MKT's? Is there a lot of wire between them?
Ray.
poynton said:
I realise that you have revised the circuit...my problem is that I cannot get hold of 10.0uH chokes !
Any chance you could compute the values for me with 4.7uH ????
Andy
Hi Andy,
I'll do my best....
What attenuation would you like the filter to have?
Ray.
6h5c said:
Hi.
I believe that HF ripple reduction is better if you use many smaller caps rather than 1 large value.
ESR's are in parallel, inductances are smaller etc.
There is also a case for putting a small inductor between the caps to form a LCR net.
Andy
